The title asks it adequately but just to be certain:
Paper day trading is frequently time-sensitive enough that if the market data delay doesn’t match the trade execution delay, the results of paper day trading are less reliable than they should be.
So it’s important to know whether paper trades are executed against current market conditions or against 15 minute delayed market conditions.
Which is it?
Paper trading simulates live trading by filling orders based upon the current non-delayed NBBO quotes. For buy orders the fill price is the current ask. For sell orders it is the current bid. The volume is not taken into account so be careful with low liquidity stocks. There is a random function to periodically create a partial fill which may result in the order filling at multiple prices. However those would still be the NBBO quote at that time. This is probably is a bit pessimistic for fill prices. Live fill prices typically see a small price improvement.
I compared about 500 buy and sell orders between paper and live trading. The difference in price was in the average range of 0.01-0.05% in favor of live trading. (paper was more pessimistic usually.) from this I can conclude that paper trading prices are in general (except for rare cases of bugs and hiccups) very similar to live trading prices.
@ishaybas Thanks for the great feedback. That is typical. There is no ‘price improvement’ in paper trading while live trading often orders are filled a bit better than the current quote.